Oligocene development of the West Antarctic Ice Sheet recorded in
eastern Ross Sea strata
Christopher C. Sorlien Institute for Crustal Studies, University of California–Santa Barbara, Santa Barbara, California 93106, USA
Bruce P. Luyendyk
Department of Earth Science, University of California–Santa Barbara, Santa Barbara, California 93106, USA
Douglas S. Wilson
Robert C. Decesari*
Louis R. Bartek Department of Geological Sciences, CB 3315, University of North Carolina, Chapel Hill, North Carolina 27599-3315, USA
John B. Diebold Lamont-Doherty Earth Observatory, Columbia University, 61 Route 9W, Palisades, New York 10964-8000, USA
ABSTRACT
Seismic-reflection data from the easternmost Ross Sea image
buried scour-and-fill troughs and flat-topped ridges interpreted as
having formed by glacial erosion and deposition during the Oligocene. The NNW-SSE orientation of the troughs and lack of similar
Oligocene glacial features within the central Ross Sea suggests that
the ice issued from the highlands of Marie Byrd Land located 100 km
away and that portions of the West Antarctic Ice Sheet formed earlier
than previously accepted. Existing global climate models (GCMs) do
not produce West Antarctic ice caps for the Oligocene, in part due
to low elevations modeled for that time. Evidence for Oligocene ice
beyond the paleocoast suggests a higher elevation for the early Cenozoic Marie Byrd Land and Ross Embayment than at present.
Keywords: Antarctic Ice Sheet, Marie Byrd Land, West Antarctica, seismic stratigraphy, Ross Sea, Oligocene, basins.
INTRODUCTION
Oxygen isotopes have been interpreted to indicate the growth of
continental-sized Antarctic ice sheets at the Eocene-Oligocene boundary (Zachos et al., 2001). Ice is modeled to have first accumulated on the
high-elevation plateau of East Antarctica (DeConto and Pollard, 2003), but
the extent to which West Antarctica participated in Oligocene glaciation is
not known. Indeed, the oldest seismic stratigraphic evidence for regional
grounded ice in the eastern Ross Sea is variously interpreted as late Oligocene (Bartek et al., 1992), middle Miocene (Bart, 2003), and late Miocene
(De Santis et al. 1999). Glacial erosion between ca. 28 Ma and 15 Ma of
dated volcanoes in the interior of northern Marie Byrd Land as well as
hydrovolcanic rocks suggest a small Oligocene ice cap there (Rocchi et al.,
2006). More data are required to determine how important West Antarctica
was to the major early Oligocene expansion of the Antarctic cryosphere.
Global climate models (GCMs) are used to understand midCenozoic ice-sheet development. GCMs incorporate atmospheric composition, paleo-elevation, heat transport by ocean currents and atmosphere,
and orbital parameters (DeConto and Pollard, 2003). Late Oligocene to
present uplift of a dome in northern Marie Byrd Land has been proposed
to explain the current elevation of a Cretaceous erosion surface there
(LeMasurier and Landis, 1996). Regional subsidence due to crustal cooling is thought to have occurred after the end of Late Cretaceous extension for much of Marie Byrd Land and the Ross Sea (Decesari, 2006;
Luyendyk et al., 2001). Subsidence means that, at least for areas outside
the dome, Marie Byrd Land was much higher-standing during Oligocene time than the mid-Cenozoic West Antarctic archipelago calculated
by DeConto and Pollard (2003). Their GCM was able to grow a midCenozoic ice sheet only on an elevated East Antarctica, with no regional
ice cap on West Antarctica. If data show that Oligocene strata, now far
below sea level, preserve evidence of West Antarctic ice, then the regional
subsidence model is supported.
*Current address: ExxonMobil, 222 Benmar Drive, Houston, Texas
77060, USA.
SETTING AND RESULTS
The West Antarctic rift system includes the Ross Embayment, the
region now covered by the Ross Sea, and the Ross Ice Shelf (Fig. 1A).
Cretaceous and early to middle Cenozoic lithospheric extension between
East and West Antarctica has produced several topographic highs and
adjacent major basins (Fig. 1A; Cooper et al., 1995). We have recently
acquired seismic-reflection and multibeam bathymetry data in areas that
were covered by Ross Ice Shelf until 2000, when giant icebergs calved
(Fig. 1B). Details of the acquisition and processing of the reflection data
are available elsewhere (Decesari, 2006). Higher-resolution single-fold
reflection data were recorded using a single air gun (Fig. 2) coincident
with deep-imaging multichannel data. Both reflection and bathymetry data
image basement exposed at or near the seafloor on a ridge extending north
from the Roosevelt Island ice dome (Figs. 1B and 3A). An ~3-km-deep
sedimentary basin, here named the Roosevelt subbasin, is present between
this ridge, named Roosevelt Ridge, and Marie Byrd Land (Fig. 3).
Observations
A succession of broad (10–20 km) and narrow (2–5 km) buried
U-shaped troughs are present in Roosevelt subbasin (Figs. 2 and 4). A
succession of flat-topped ridges separates the narrow troughs. All of the
troughs and flat-topped ridges have similar amplitudes, ~150 m, and are
between 700 and 900 m depth, with the top of the succession of troughs
between 0 and 150 m below seafloor (Fig. 2). Below the interval of the
troughs is a 200-m-thick succession of offlapping sequences that have an
apparent (two-dimensional) progradation toward the southwest (between
cyan and violet reflections on Fig. 4B). These sequences are associated
with a flat-topped ridge at their distal (southwest) end, at 1200–1400 m
depth (Fig. 4C). The broad troughs can be individually correlated between
four parallel profiles and are aligned NNW-SSE, and our tentative correlations between profiles indicate that the narrow troughs and flat-topped
ridges may be aligned NW-SE to W-E (Fig. 1B).
Correlation of Chronostratigraphic Control
We use the Antarctic Offshore Acoustic Stratigraphy (ANTOSTRAT;
Cooper et al., 1995) stratigraphic nomenclature for sediments filling the
Eastern Basin, as modified by Luyendyk et al. (2001) (Fig. 3). In the Eastern
Basin, the units are named “RSS-” sequences (RSS-1 = oldest) separated
by regional unconformities “RSU-” (RSU1 = youngest). These units and
unconformities were dated by Deep Sea Drilling Project (DSDP) Leg 28,
which sampled the west flank of the Eastern Basin (Fig. 1A; Hayes and
Frakes, 1975). The oldest marine sedimentary rocks DSDP recovered are
dated as late Oligocene (RSS2-upper; ca. 25 Ma), based on foraminifera
from Site 270 (Leckie and Webb, 1986). Although the correlation path from
DSDP Site 270 to the northeast Roosevelt subbasin is 450 km long, the correlation is straightforward because of uniform marine deposition and near
total lack of deformation. The underlying RSS2-lower strata, including the
offlapping sequences, have not been sampled, but they are also presumed
to be Oligocene because they overlie the putative Oligocene RSU-6 unconformity (Fig. 4B; Cooper et al., 1995; Luyendyk et al., 2001). The RSS2upper reflections seen in the northern part of the survey (Fig. 4D) are from
© 2007 The Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected].
GEOLOGY,
May
2007
Geology,
May
2007;
v. 35; no. 5; p. 467–470; doi: 10.1130/G23387A.1; 4 figures.
467
es
W
04
20
M
tic
tarc
nsan
°E
Tra
50
Roosevelt
Island
)
(1B
Marie
Byrd
Land
F-
E--E
D
C--C--
ins
--
C
-
B--B--B
0.8
Two-Way Travel Time (s)
°S
-D
-
1
80
m
-500
160°W
ca
cti
4c
N
Rockefeller
Mtns
Kie
10 km
Prestrud
Inlet
Figure 1. A: Gray shade shows major Ross Sea basins with south
edge along modern ice shelf edge, existing drill sites (Sites 270–274
from Deep Sea Drilling Project [DSDP] Leg 28), and Roosevelt subbasin (RvSB). Dashed curves B–F show ice streams. Bold dashed
line shows seismic-reflection profiles used for age constraints
(Cooper et al., 1995; Luyendyk et al., 2001). B: Locations of cross
sections and seismic profiles (Figs. 2 and 4), parts of Marie Byrd
Land, depths and elevations of top of rock (shaded), grounding line
(heavy gray curves), and edges of Ross Ice Shelf in 1987 and 2004.
Inset map gives axes of troughs as thick curves, edges of troughs as
thin curves, and flat-topped ridges as dashed thin curves; seismic
track lines (gray) are also shown.
NBP-0301-27
Southwest
5 km
Northeast
Single channel
26x vertical exaggeration at seafloor
RSU4
+4a+5?
1.0
1.2
Moraines
the base of that interval and underlie the ca. 25 Ma strata cored at DSDP Site
270. Basal RSS2-upper along with RSS2-lower thus predate 25 Ma.
Unlike along the correlation path shown on Figure 1A, the troughs
and flat-topped ridges cannot be dated by direct correlation from northeast to the central Roosevelt subbasin because the sedimentary sequences
containing them pinch out along the southern part of the profile shown
in Figure 4D. However, based on their seismic character and similarity
between their occurrence southwest of point T1 and the occurrence of
RSS-2-upper north of point T2 (Fig. 4), a correlation spanning ~30 km, we
interpret these units to be similar in age to the late Oligocene basal part of
RSS-2-upper. Additionally, along the eastern margin of the Eastern Basin,
the early Miocene part of RSS-2 and younger sequences pinch out and do
not occur south of ~77°S (Luyendyk et al., 2001). Unconformities above
RSS-2 dip northward throughout the Eastern Basin (De Santis et al., 1999;
Luyendyk et al., 2001), so post-Oligocene units are unlikely to be present
at 800 m depth in Roosevelt subbasin.
DISCUSSION AND CONCLUSIONS
We interpret offlapping sequences (Fig. 4B) to be deltas formed
adjacent to the grounding line of a glacier (see Bart, 2003) and the synchronous flat-topped ridge (Fig. 4C) to be a moraine. Above these deltas,
we interpret the flat-topped ridges to be buried moraines and the broad
troughs to have been carved by glaciers. If the flat-topped ridges were
instead erosional remnants, their internal reflections would be horizontal.
The internal reflections are instead subparallel to the edges of the troughs,
468
+250 m
RvSB
Ross
Ice
Shelf
Figure 2. Nonmigrated single-channel seismic profile NBP 0301-27
(Fig. 1B). We interpret broad
U-shaped troughs to have been
successively cut and filled by ice.
Flat-topped ridges are interpreted
as moraines deposited in successive stages. This profile coincides
with upper part of depth section in
Figure 3B. Separate sequence
boundaries within middle of Eastern Basin (Fig. 3A) merge eastward
and are labeled RSU4+4a+5.
4d
cier
ca
cti
ar
nt
50 km
00 m
-10
Eastern
Basin
Fig. 3a
ou
nta
77
°S
l Gla
ar
nt
2 4b
75°S
271
272
270
1987
tA
A
st
gh
ou
Tr
Ross
Sea
273
78
°S
B)
15
5°
W
Pacific
Ocean
Ea
e
ar
180°
Ad
70°S
16
0°
W
A)
274
Successive cut-and-fill scours
as would be expected for a moraine (Fig. 2). Diapirism is not a likely explanation for the ridges because trough fill onlaps them without upwarping
on Figure 2 and nearby. Continuous reflections cross beneath the ridges
in Figure 2 and at deeper levels on both the coincident multichannel stack
and a nearby stack, with slight pull-up beneath each ridge interpreted as a
velocity effect. The sizes of the features are not unexpected for a moraine:
a series of moraines is imaged on seismic-reflection data in Glacier Bay,
Alaska, that has up to 200 m amplitude and a spatial pattern that is similar
to that in Figure 2 (Gulick et al., 2004). We do not expect that a series of
10–20-km-wide troughs, with consistent ~150 m relief, could be carved
by nonglacial processes on the shelf of a passive margin.
A fluvial explanation for the troughs and ridges is precluded if they
formed below sea level, and the expected late Oligocene water depth gives
a minimum thickness for grounded ice. The depositional load of RSS-2
may be a factor in tilting and subsidence, but younger units of sufficient
thickness to cause post-RSU5 subsidence are not present in close proximity to Roosevelt subbasin (Fig. 3A; Luyendyk et al., 2001). Thermal subsidence (e.g., McKenzie, 1978) resulting from earlier extension is a viable
mechanism for moderate subsidence, but reasonable stretching factors and
extension ages from Late Cretaceous to early Tertiary limit post–25 Ma
thermal subsidence to at most ~400 m (Decesari, 2006). The current
700–900 m depth of the trough and fill sequences indicates that they were
deposited in marine conditions at depths of several hundred meters.
The troughs were filled with sediment as ice retreated; then, as ice
advanced, a new trough was cut into the fill (Fig. 2). These successive
GEOLOGY, May 2007
170 W
175 W
165 W
B) Southwest 0301-27 Northeast Miocene?
5 km
160 W
100 km
-0.5
Roosevelt
sub-basin
RSS-5
-2.5
Land
Eastern
Basin
RSS-2
Olig-Miocene
RSS-1
Early(?)
Oligocene
RSS2-lower
RSU6
Eocene?
RSS1-upper
2
Basement
-3.0
cyan
ent
em
bas
near
Roosevelt
Island
RSS-3
Lower Miocene
Marie B
yrd
Depth (km)
igh
ral H
Cent
-2.0
RSS2-lower
S
SS
1
RSS-4
Lower-Mid Miocene
-1.0
-1.5
RSU4+
4A+5
Late
Oligocene
violet
RSS-7&8
Plio-Pleistocene
0
Depth (km)
A) 180
RSU7
RS
S1
-l
Cretaceous?
ow
er
40x VE
Figure 3. Both cross sections are drawn from interpreted depth-migrated seismic-reflection profiles. A: Cross section across eastern Ross
Sea at ice-shelf edge (Fig. 1A) showing that strata in Eastern Basin and Roosevelt subbasin are separated by a basement high. B: Detail of
cross section across western Roosevelt subbasin. Colors of horizons are same as in Figure 4. Thin black lines within RSS2-upper outline
flat-topped ridges seen in Figure 2. VE—vertical exaggeration.
B Southwest
25
X
18T1
lower
RSS2-
U6
er
1-upp
RSS
-17
0
500 m
North
0301-17B
D
Z
bend
2
2 km ion
ject
pro
5 km
20x V.E.
T2
2x V.E.
X
South
tie 18
0301-25A
Y
RSS2-upper
RSS2-lower
Depth (m)
1100
X
er
RSS2-low
w 301
C
Northeast tie 25A
T1
1
RSS2-lower
Depth (km)
-78 15'
01
0301-18
1
01-
03
10x V.E.
Red
T2
Y03
-78
5 km
bend
0301-25
-77 45'
-10
9601
Z
w
-160
-161
Depth (km)
A
RSS1-upper
U6
U7
RSS1-lower
1500
M
M
M
2
Figure 4. A: Location map for profiles and reference points. B: Migrated multichannel seismic-reflection depth section across eastern half
of Roosevelt subbasin to its intersection with profile in D at (X) (Figs. 1B and 4A). Offlapping sequences (between cyan and violet) are interpreted as deltas deposited near grounding line of a glacier. These features are within lower part of RSS-2-lower (pre-25 Ma), beneath a 20-kmwide U-shaped trough. C: Migrated depth profile of flat-topped ridge interpreted as moraine and as synchronous with distal part of offlapping
sequences. Profile is located in A, and its projected location is indicated by dashed arrow and circle in B. D: Nonmigrated single-channel
seismic-reflection profile displayed in depth, shown between its intersections with profile in B (X) and NBP 9601 Line 10 (Z) from Luyendyk
et al. (2001) (also http://projects.crustal.ucsb.edu/colbeck/Colbeck5a.html). Line 10 is directly correlated to Deep Sea Drilling Project (DSDP)
Site 270 along dashed path in Figure 1A. Trough at T2 is overlain by basal RSS2-upper, so it predates 25 Ma. We interpret that troughs seen
on Figure 4B are the same age. V.E.—vertical exaggeration.
GEOLOGY, May 2007
469
advances and retreats are consistent with the dynamic Antarctic ice caps
interpreted by Pekar et al. (2006) for the late Oligocene. The lower offlapping sequences are restricted to the eastern half of Roosevelt subbasin,
adjacent to Marie Byrd Land, and may represent the first appearance of
glaciers in the present area of the eastern Ross Sea. Ice may have been
more extensive in paleohighlands located to the southeast (Fig. 1B).
We are confident that the troughs and flat-topped ridges are of glacial origin, but because the interval that contains them pinches out along
Figure 4D, the interpretation that they predate 25 Ma is based on a comparison of seismic stratigraphy (Figs. 4B and 4D). The older offlapping
sequences can be continuously correlated to beneath pre–25 Ma basal
RSS2-upper strata (Figs. 4B and 4D). It is, however, less certain that these
offlapping sequences were deposited proximal to a glacier. Our preferred
interpretation is that all of the troughs, ridges, mounds, and offlapping
sequences predate 25 Ma and postdate RSU6. We infer RSU6 to have
formed during sea-level falls of ~50 m between 33.7 and 33.5 Ma and/or
between 28.3 and 28.1 Ma (Miller et al., 2005).
Narrow erosional troughs and reflective mounds or ridges are imaged
within pre–25 Ma strata on the west flank of Roosevelt Ridge, but such
features are not present in the deep part of the Eastern Basin near the iceshelf edge, even though Oligocene sedimentary rocks are preserved there.
If the ice that scoured the troughs in Roosevelt subbasin came from East
Antarctica, then similar features related to glacial erosion and deposition
might be preserved in the Eastern Basin near the ice-shelf edge, and they are
not seen. These observations suggest that the ice within Roosevelt subbasin
originated from West Antarctica and that these sedimentary sequences may
contain a direct record of the history of the West Antarctic Ice Sheet.
Incision of troughs hundreds of meters below sea level by a grounded
ice sheet requires substantial ice on a nearby landmass. Presently, the
nearest high-elevation rock surface is in the Rockefeller Mountains; these
highlands of Marie Byrd Land are located over 100 km to the southeast
(Fig. 1B). The modern Prestrud Inlet (Fig. 1B) is a flow path for ice
sourced from highlands even farther southeast. Evidence for a late Oligocene ice cap in northern Marie Byrd Land (Rocchi et al., 2006) suggests
the possibility of a regional, late Oligocene West Antarctic Ice Sheet.
If the seismic stratigraphic correlation (Fig. 4) presented here is correct,
then this paper documents the first evidence for marine Oligocene grounded
ice located far from an elevated source in Marie Byrd Land. This places the
initiation of the West Antarctic Ice Sheet during the late Oligocene and helps
explain global ice volumes calculated for that time as high as 125% of the
modern East Antarctic Ice Sheet (Pekar et al., 2006). Input of a high-standing
Oligocene Marie Byrd Land, rather than an archipelago, into a GCM, favors
growth of the West Antarctic Ice Sheet for expected Oligocene atmospheric
composition and heat transport. Even if northern Marie Byrd Land dome
experienced Oligocene and Neogene rock uplift and surface uplift (Rocchi
et al., 2006), this does not preclude regional subsidence of the Roosevelt
subbasin and adjoining western Marie Byrd Land, because the two regions
are spatially distinct. If the deeper offlapping sequences and associated
flat-topped ridge are indeed glacial proximal deltas and moraines, then an
episode of Oligocene grounded ice that predates formation of the troughs
imaged in Figure 2 occurred near or beyond the paleocoast of Marie Byrd
Land. It is possible that Oligocene grounded ice only affected the SE corner
of the Ross Sea, adjacent to Marie Byrd Land, which is not inconsistent with
Miocene initiation of grounded ice in the area of the Eastern Basin some
hundreds of kilometers to the northwest (Bart, 2003; De Santis et al., 1999),
or with late Oligocene grounded ice there sourced from East Antarctica or
the Central High (Bartek et al., 1992).
ACKNOWLEDGMENTS
We thank the captain and crew, Raytheon staff, and the scientific party of
the US RVIB Nathaniel B Palmer during expeditions 03-01 and 03-06. Lloyd
Burckle, James Kennett, and Steve Pekar provided helpful comments. Reviews by
Steve Pekar, Wesley LeMasurier, and Sergio Rocchi are appreciated. Dan Herold
and Matt Ralston of Parallel Geophysics Corporation provided valuable help with
470
seismic processing. Seismic Micro Technology Corporation donated the Kingdom
Suite software. This work was supported by the National Science Foundation (NSF)
grants OPP-0088143 to Luyendyk and Wilson, OPP-0087392 to Bartek, and OPP0087983 to Diebold. This is contribution 781 of the Institute for Crustal Studies.
REFERENCES CITED
Bart, P.J., 2003, Were West Antarctic Ice Sheet grounding events in Ross Sea a
consequence of East Antarctic Ice Sheet expansion during the middle Miocene?: Earth and Planetary Science Letters, v. 216, p. 93–107, doi: 10.1016/
S0012–821X(03)00509–0.
Bartek, L.R., Sloan, L.C., Anderson, J.B., and Ross, M.I., 1992, Evidence from
the Antarctic continental margin of late Paleogene ice sheets: A manifestation of plate reorganization and synchronous changes in atmospheric circulation over the emerging Southern Ocean?, in Prothero, D.R., and Berggren,
W.A., eds., Eocene-Oligocene Climatic and Biotic Evolution: Princeton,
New Jersey, Princeton University Press, p. 131–159.
Cooper, A.K., Barker, P.F., and Brancolini, G., eds., 1995, Geology and seismic
stratigraphy of the Antarctic margin: American Geophysical Union Antarctic Research Series, v. 68, 301 p., atlas, CD-ROM.
Decesari, R.C., 2006, The Mesozoic and Cenozoic depositional, structural, and
tectonic evolution of the Ross Sea, Antarctica [Ph.D. thesis]: Santa Barbara,
University of California, 232 p.
DeConto, R.M., and Pollard, D., 2003, A coupled climate-ice sheet modeling
approach to the early Cenozoic history of the Antarctic ice sheet: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 198, p. 39–52, doi:
10.1016/S0031–0182(03)00393–6.
De Santis, L., Prato, S., Brancolini, G., Lovo, M., and Torelli, L., 1999, The Eastern Ross Sea continental shelf during the Cenozoic: Implications for the
West Antarctic Ice Sheet development: Global and Planetary Change, v. 23,
p. 173–196, doi: 10.1016/S0921–8181(99)00056–9.
Gulick, S.S., Powell, R.D., Jaeger, J.M., Cowan, E.A., Mayer, L.A., Mix, A.C.,
Finney, B.P., Pisias, N.G., Prahl, F., and Stoner, J.S., 2004, Glacial advances
and retreats in tectonic southeast Alaska during the Little Ice Age and Last
Glacial Maximum: Preliminary results from EW0408 [abs.]: Eos (Transactions, American Geophysical Union), v. 85, no. 47, Fall Meeting supplement, abstract H51A–1099.
Hayes, D.E., and Frakes, L.A., 1975, General synthesis, Deep Sea Drilling Project Leg 28, in Hayes, D.E., and Frakes, L.A., et al., Initial Reports of the
Deep Sea Drilling Project, Leg 28: Washington, D.C., U.S. Government
Printing Office, p. 919–942.
Leckie, R.M., and Webb, P.-N., 1986, Late Paleogene and early Neogene foraminifers
of Deep Sea Drilling Project Site 270, Ross Sea, Antarctica, in Kennett, J.P., and
von der Borch, C.C., et al., Initial Reports of the Deep Sea Drilling Project, Volume 90: Washington, D.C., U.S. Government Printing Office, p. 1093–1120.
LeMasurier, W.E., and Landis, C.A., 1996, Mantle-plume activity recorded by
low-relief erosion surfaces in West Antarctica and New Zealand: Geological
Society of America Bulletin, v. 108, p. 1450–1466.
Luyendyk, B.P., Sorlien, C.C., Wilson, D.S., Bartek, L.R., and Siddoway, C.H.,
2001, Structural and tectonic evolution of the Ross Sea rift in the Cape
Colbeck region, Eastern Ross Sea, Antarctica: Tectonics, v. 20, p. 933–958,
doi: 10.1029/2000TC001260.
McKenzie, D., 1978, Some remarks on the development of sedimentary basins:
Earth and Planetary Science Letters, v. 40, p. 25–32, doi: 10.1016/0012–
821X(78)90071–7.
Miller, K.G., Kominz, M.A., Browning, J.V., Wright, J.D., Mountain, G.S., Katz,
M.E., Sugarman, P.J., Cramer, B.S., Christie-Blick, N., and Pekar, S.F.,
2005, The Phanerozoic record of global sea-level change: Science, v. 310,
p. 1293–1298, doi: 10.1126/science.1116412.
Pekar, S.F., Harwood, D., and DeConto, R., 2006, Resolving a late Oligocene
conundrum: Deep-sea warming versus Antarctic glaciation: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 231, p. 29–40, doi: 10.1016/
j.palaeo.2005.07.024.
Rocchi, S., LeMasurier, W.E., and Di Vincenzo, G., 2006, Oligocene to Holocene
erosion and glacial history in Marie Byrd Land, West Antarctica, inferred
from exhumation of the Dorrel Rock intrusive complex and from volcano
morphologies: Geological Society of America Bulletin, v. 118, p. 991–
1005, doi: 10.1130/B25675.1.
Zachos, J., Pagani, M., Sloan, L., Thomas, E., and Billups, K., 2001, Trends,
rhythms, and aberrations in global climate 65 Ma to present: Science,
v. 292, p. 686–693, doi: 10.1126/science.1059412.
Manuscript received 21 September 2006
Revised manuscript received 22 December 2006
Manuscript accepted 29 December 2006
Printed in USA
GEOLOGY, May 2007